title esr studies of the hexagonal hydrate of molybdenum oxide … · 2016. 6. 15. · title esr...
TRANSCRIPT
Title ESR studies of the hexagonal hydrate of molybdenum oxideand its changes on treatments in vacuo and in hydrogen
Author(s) Sotani, Noriyuki
Citation The Review of Physical Chemistry of Japan (1976), 46(1): 1-8
Issue Date 1976-06-30
URL http://hdl.handle.net/2433/47023
Right
Type Departmental Bulletin Paper
Textversion publisher
Kyoto University
The Review of Physical Chemistry of Japan Vol. 46 No. 1 (1976)
THE REVIEW OP PHYSICAL CHEMISTRY OF JAPAX, VOL. 46, Nn.l, l9%6
ESR STUDIES OF THE HEXAGONAL HYDRATE OF MOLYBDENUM OXIDE
AND ITS CHANGES ON TREATMENTS
IN VACUO AND IN HYDROGEN
BY 1URIYCRI SOTA\I
The hexagonal hydrate of molybdenum trioxide and iu change on treatmenu in vacuo and in hydrogen were studied by ESR spectroscopy. The hydrate gave no ESR signal, but the signal due to No(V') appeared in the case of the reduction of hexagonal b1oOs and rhombic AfoOs. >lo(V) was Formed io oxygen dehcient \IoOI as a defect during the reduction. The signal intensity changed according to the degree of the reduction. The intensity of the reduced sample decreased when oxygen was added and recovered its value by evacuation. This means that Oy molecule adsorbs weakly on \lo(V). The mechanism of the change is intensity by oxygen adsorption and desorp[ion is assumed as follows:
Oz Oz
-Mo (V)-O-Mo (V)- -~ -Mo(VI)-O-Mo(IV)-
or Oz Oz
-+ -Mo (VI)-~-MO (VQ-
Introduction
There are many ESR works of ~JoOaLZ), \4o0a supported on AIz0a3-'} or SiQs•]1 and molybdate
catalyst%'ii7, in order to know the correlation between \Jo(V) in molybdenum oxide and iu catalytic
(Received January 2i, 1976) 1) P. F. Cornaz, J. H. C. van Hooi, F. J. Plujin and G. C. .4. Schuit, Disaus. Faraday, Sx., dI, 290 (1966)
2) h1. ]. Jlasson and M. J. \echtscheim, Bull. Soc. Ckim. France, 3933 (1968} 3) G. Ii. Boleskov, B, A. llzisylo, V. Dl. Emelynov and V. B. 6azansky, Dak1. dead. .1'auk SSSR, 150,
829 (1963) 4) V. ]shii and f. Matsuura, .\'ippm+ Kagakn Xaftki, 89, 553 (1968) 5) M. Ak7moto and G. Echigoya, Chem. Lelt., 305 ;1971) fi) K. 5. Seshadri, F. E. Nassoth and L. Petrakis. !: Caml.. 19, 95 (1970)
7) K. AI. Saucier, T. I)ozono and R. N'ise, ibid., 23, 270 (1971) 8) J. M. Peacock. M. J. Sharp, A. J. Parker, P. G. 9shmore and J. A. Rocky, ibid., Ig, 379 (1969)
9) K, S Seshadri and L. Petrakis, J. Pkys. ChenL, 74, 4702 (1970) 10) 'C, Rwan and hf, Sato, .Fippan /Cagaka b"aisl+i, 91, 1103 (19]0) I l) H. Ueda end N. Todo, Buq. Ckem. Sot. Jap.; 43, 3698 (1970)
The Review of Physical Chemistry of Japan Vol. 46 No. 1 (1976)
2 '~. Sotani
activity. The signals, which were assigned to Mo(V), were observed on supported molybdenum
oxides'71 and binary catalystl-sy, but, on the unsupported molybdenum oxide, any signal due to Mo(V)
was not observed at room temperatureL21. Therefore. it was assumed that the supporters or the con-
stituents in binary catalysts stabilized the Mo(V) state. However, Cornaz et aLrl detected the signal
due to Mo(V) at -195°C on single MoOa sample.
The author[''-1 obtained the hexagonal hydrate of molybdenum trioxide, a new form of the hydrate
of molybdenum trioxide, in his studies on the molybdenum oxides and the hydrates. The properties
and the behaviors of the hexagonal hydrate on the treatments in vaeero, in hydrogen and in the
hydrogen-[hinphene mixture a[ various temperatures were studied by means of \-rap diffraction
analysis, IR spectroscopy and DTA-TG a. Inthese studies, the author made clear the compositional
and structural changes of [he hexagonal hydrate due to dehydration, transformation, reduction and
sulfurization, accompanied by these treatments.
In these reactions, \io(V) is expected to be formed, even though any evidence of hIa(V) com-
pounds was not obtained. In connection with the previous work[''-1, the author studied. by ESR spec-[roscopy, [he sample and its change of the state resulted from the treatments in vacuo, in air and in
hydrogen, rvi[h [he purpose of revealfn_¢ the role of bio(V) in the change of the state of the hydrate.
Experimental
Sample
The hexagonal hydrate of molybdenum trioxide was prepared by the same method as has been
previously describedta>.
ESR measurements
The ESR measurements mere performed ona JES 3B% Type (JAPAN ELECTRON OPTICS
LA[i.) spectrophotometer operating at 9.4 SJIHz. The magnetic field w•as modulated by 100KHz, and
the first derivative of [he spectrum was recorded. The peak-to-peak modulation amplitudes from 3 Co
5 gauss were used in recording the spectrum.
The sample tubes were made of quartz of S mm is o. d. and 4 mm in i. d. The tube had a stop cock
and could be evacuated by connecting it to a conventional vacuum system with a ground glass joint.
The samples were subjected to ESR analysis at room temperature after evacuation for 30 min under a
vacuum of 10-'-10-6 tort. The amounts of the sample used for ESR measurements were 4 to 40 mg.
Within [his range, the signal intensity increased in exact proportion to the sample weight. Intensities
were calibrated by using one of the intensity of the lines from ili n~' sample.
X-ray studies
I2) N, Sotani, iGid., 48, 1820 ([975) 13) N. Sotani, Y. 5aito, D7. Oita and U. Hasegawa, :Vippon Xagaku %aisM, 673 (1974)
The Review of Physical Chemistry of Japan Vol. 46 No. 1 (1976)
ESR Studies of [he Hexagonal Hydrate of \iolybdenum Oxide 3
The precise method of Xaay diffraction analysis was described previouslyta).
Experimental procedure
(1) Heut treatment in vaaro. The hydrate sample was Created at 350' and 400-C under the vacuum of 10-'-10-' tort for 3O min.
(1) Reduction. Reduction experiments were performed both in the closed system and in the flow =_ystem. In [he experiments in the closed system, the hydrate sample was packed in the ample tube.
After the evacuation at room temperature for 3O min under the vacuum of ]0-'-30-s tort, the sample
was reduced by the hydrogen purified by passing through the platinum catalyst and molecular sieves.
.after the reduction. the sample was subjected to ESR analysis. On the other hand. the hydrate sample
was reduced in the flow system by atmospheric hydrogen with the rate of 0.1 mot/hr. After the re-
duction. [he sample was quickly transferred to the tube for ESR measurement. evacuated and subjected
to ESR analysis
Results
Treatment in racuo
The hexagonal hydrate of molybdenum trioxide gave no ESR signal in air or in vacuum of 10-1-
10-' tort at room temperature. Pig. 1 shows the spectra of the sample treated fn vacuo for 30 min at
3iO` and 4O0'C. The signal with g=1.934 and g=2.001 was obsen~ed when the sample was treated at
35O'C.. though any structural change was not observed by X-ray diffraction analysis. It was reportedtz)
that the dehydration of [he hydrate took place within 5 min at 350°C and the hydrate transformed to
anhydrous hexagonal 1IoO~ (denoted a; (H)) by heating, without any change in structure. By treatment
at 400°C. the g•values of the signal shifted to g=1.911 and g=1.991 (Fig. l). Ey the X-ray studies,
the sample was transformed to rhombic \foO, (denoted as (R)).
1.934
t 1.001
~1-915 "`J h _ 1.995 A 1.929
T-3i0°C
Fig, t The E3R spectra of the sample (.4; T=330'C in ream, B ~ T-400'C in cacao, C; reduced with hy-
T=400'C drogen at 330'C for 30 min)
1.003
100 i H gauss
1.Si
30 min
The Review of Physical Chemistry of Japan Vol. 46 No. 1 (1976)
4 K. So[ani
Reduction of the sample
The hydrate sample was reduced by hydrogen of 28-31 tort a[ 35D-C in the closed system and was
subjected to ESR analysis after ecatuation. The signah similar to that of the sample obtained by heat
treatment in vacno, appeared by the 3 min reduction. Fig. 1 shows the spectrum of the sample reduced
for 30 min. The line shape of the spectrum is asymmetric with a main signal with the g-value near L93
and a shoulder with g=1.87, showing the line shape of the nonoriented polycrystalliaes. The g-value
of the main signal Changes from 1.936 to 1.929 and the intensity increases with the progress of the
reductio¢.
A small signal with g=2.003 was also observed.
According to X-ray diffraction analysis, the hydrate sample changed as follows: (HN(R~.MoO;
(denoted as (D)). when it was reduced with hydrogen at 350°C in the Now system. The composition of the reduced sample was a mixture of (H), (R) and (D), but the relative amounts of them varied
according to the degree of the reduction. However, the reduction at 350-C was so slow that the hydrate
sample was reduced with an atmospheric hydrogen at 400°C to obtain the further reduced sample. The
composition of the sample was. at first, a mixture of (fi), (R) and (D). By the 40min reduction the
composition changed to a mixture of (R) and (D). and by the 60 min reduction, most of (ll), the final
reduction product. The oxides of Mo(Y) and blo(III) states n-ere not detected by X-ray diffraction
analysis throughout these reduction experiments. The ESR signal with g=1.925; and g=1.996 appeared in the 5 min reduction. The signal intensity
of g=1.925; changed with the degree of the reduction. The intensity is plotted against the fraction of
(D)in the sample as shown in Fig. 2. The i¢tensity has a maximum at about 5 °b of (D) and gradually decreases to a very small slue at about 6095 of (D). The signal with g=1.996 also takes high inten-
sity at low (D) content, but Che intensity is not completely parallel [o the main signal (Pig. 2).
0.1
0 t
O T ~. .V
C .~
m y
Q
_~,•
~ f ~' :',i
0.01
0.005
0.2 0.4 D.6 0.8 Fraction of \fo0r
01.0
Fig. 2 The change in intensity of the fraction of MoOi in [he sample reduced with hydre-
gen (~: g~ 1.93, G: g-2.00)
Change in intensity of spectra of [he reduced sample with the atmosphere
Fig. 3 shows the Change in the intensities of [he main signal (Curves A. B and C) and the signal
with g=2.00 (Curves B' and C') with time, when the various atmospheres were added to the reduced
sample. The curve A shows the change in intensity when the sample was evacuated immediately after
The Review of Physical Chemistry of Japan Vol. 46 No. 1 (1976)
ESR Studies of the Hexagonal Hydrate of Efolybdenum Oxide
reduction. No change is observed. The curve B is the result when air was introduced after reduction.
Intensity decreases to a certain minimum value by [he 40 min exposure. The curve C shows the change
in intensity when the sample was stood in reduction atmosphere without evacuafion after the reduction
and, then, treated as indicated in the figure. The first decrease seems to be due to the adsorption of
H;;O produced by the reduction. The curve C also shows that the decrease and the inaese in intensity
by addition of air and by evacuation are reversible.
t.z
i.o
os
os
0.4
0 E n
0.2
a a
-e-~ (a)
0~ air
"'pa
aic0
(B)
.air
~C
~•(B')
~c~
\o0
air
evacuation
;r o 18 hr it hrJ f--
Fig. 3
_ o ., ,.
lfi Y air
\~C~) evacuation :~~ V~
hr
./air
0 110 hr
0 120 160
The change in the signal
intensity of the sample at
various atmospheres
40 80 Time mint
The signal with g=2.00 (Curves B' and C') changes parallel to the main signal Any new signal
attrihutable to adsorbed oxygen was no[ detected by addition of air.
Fig. 4 shows the change in the intensity of the signal, when the water vapor (18.G torr) was added
to the reduced sample. The intensity of the signal with g=1.93 rapidly decreases by the S min ex-
posure and keeps a constant value for further exposure. At the same time. the signal with g=2.00 disappears by the 6 min exposure.
1.0
0 ~.8 ~.
° 0.6 T
e ° 0.4 c
0.2ti c m
0 10 10 Time(
30 min)
40 50
Fig. 4 The cha¢ge i¢ the sig¢al i¢tensitp of the samp)e on [he adsorption of mater (~; reduced for 5 min, p; reduced for lO roi¢)
The Review of Physical Chemistry of Japan Vol. 46 No. 1 (1976)
6 K. Sotani
Discussion
Molybdenum can exist in any of four valences 6t. St, 4t and 3t. Hexavalent is diamagnetic
and, therefore, gives no ESR signal Tetravalent mohbdenum, even if it were paramagnetic, was re-
ported to be obsercedonly at eery low temperaturet4). Kihlborg reported)=) that no phase corresponding to [he oxide, bioOl,s, was observed, even though MoOa had been reduced by hydrogen a[ SOO-C for a long period. In the present work, the final reduction state determined by X-ray diffraction analysis
was (D) and the reduction temperature was lower than Kihlborg's reduction temperature. There[ore.
the formation of Mo(III) can not be expected.
As for \Io(V), ESR spectra with g=1.9i71ss and g=L950t7) were obsen•ed [or (V H~)_bioOCl; and
Mo0C1 -ion in HCl solution, respec[ively.:1lso for the supported molybdenum catalyst, the signal
with g-values behveen.1.924 and L950 were reported by many workers and were attributed to Mo
(V)r'a, 7.s.lo. u.ra, r7), Consideringthese precious workait is reasonable to conclude [ha[[he main signal with the g-values near 1.93. obtained in this work, is attributed to ilfo(V).
As described above, [be samples treated in .atuo or reduced by hydrogen avere composed of one
or [be mixture o((H), (R) and (D), and no other oxide was not detected. This fact shows Chat Mo(V)
is contained in (H), in (R) or in (D). The signal intensity due to JIo(\•) in the reduced sample shows
a maximum at a few percent of (D) and decreases steeply with the progress of the reduction and reaches
a very small value of about 60% of (D) as shown in Fig. 2. The above result shows that [he signal is
attributed [o Mo(V) Contained in (H) or in (R) phase.
It is interesting [hat the samples reduced to (D) by about 60% or more show only a very n•eak
ESR adsorption due to \io(\'). (D) is well known to be a p-type semicondudorta ls), These facts mean
that Che equilibrium concentrations of positive hole and of Mo(V) in (D) are not large. The fact that
the signal intensity due [o ]fo(V) takes high value at very low (D) Content seems to show that Mo(V)
is formed as a super saturated state in (H) and in (R) at the early stage of the reduction, probably be-
cause of difficulty of the deposition as (D). The different results obtained by the previous workst,z)
from the author's that MgrV) n•as formed in MoOa probably came from the different reduction con-
ditions.
~Io(V) in (H) or in (R) is formed by the following equations, aaompanied by the formation of
oxygen deficiencies or of the F-center:
2Mo(VI)tbO~-tH+
2Mo(V)tQ(O~ )t5O'-tHzO (1) at
- Mo(VI ItMO(V)t~(O~-)~-5 O'-tHzO (2)
li) 5. Ramaseshan and Suryan, Phys. Rea., 8d, i93 (19i 1) li) L. Kihlborg, Aeta Chem. Sraad., 13, 934 (1959)
16) C. R. Hare, Ivan Beranl and H. II. Gray, Iaorg. Chem., 1, 831 (1961) 17) W. M. Abraham, ]. P. Abrita, EI. E. Fogrio and E. Pasquini. !. Clrenr. Phys. Rev.. R4, 2069 (1966)
18) K. Hauffe e! al. Z. Phyr, Chem.. 196, 160,.435 (1950) 19) Friedrith and Sittig, Z. armrg. Chem., 145, 137 Q92i)
The Review of Physical Chemistry of Japan Vol. 46 No. 1 (1976)
ESR Studies of the Hexagonal Hydrate of Molybdenum Oxide 7
giving the n-type semiconductor Slahelin e! al.m> and Kawaguchill> have reported that (R) is [be n-type semiconductor. A[ [he same time with the formation of Mo(V), the formation of the F-center
is expected. according to the equation (2). The g-value of the F-realer is expected to have a g-value
close to that of free electron value of 2.0023, as known from the values obtained in alkali halides~~~>. The small signal with g=2.003 in the ;ample heated in vaaro and fn hydrogen is assumed [o he attri-
buted to the F-center. CornazU also observed the signal with g=1.943 and 2.001 at -195'C, when
Moos w'as reduted by butene or treated under a vacuum of 10"' torr, and proposed that the signal with
g=2.001 was attrihuted to the F-center. The measurements of the changes in intensity of the reduced sample (Fig. 3) show that Afo(V),
produced by the reduction. disappeazs on adsorption of oxygen (Curse B and C) or water (Fig. 4), but it is reproducibly recovered by the evacuation a[ room temperature. The other part of Mo(\'), however,
fs left unchanged by oxygen adsorption. The above facts show that a part of ~Io(V) is situated in the
bulk, and the other part on the surface or in the layer close to the surface disappears by oxygen ad-
sorption. The Reenter is also situated on the surface or in [he layer close to the surface and disappears
by oxygen adsorption. The reversibility of the changes in intensity shows that [he deficiencies in the
crystal remain unaffected by oxygen adsorption and, at the same time, shows that the adsorption of
oxygen is very weak. IC the disappearance of \1o(V) by the adsorption of oxygen resulted from the electron transfer
from adsorbed oxygen to IDfo(V), the formation of 0-~'~>, Os ~•~1, Oa ~m3, Oz~-1) or 0°-" is ex-
pected. O", OQ or Os' must show a new signal, but such a new adsorption did not appear in ESR spectra. Kazansky et al.ai> reported [hat the signal due to Mo(V) in 51o0a supported on SiOs decreased by oxygen adsorption and new triplet signal with ga=2.018, g~=2.010 and ga=2.004 was observed and
assigned it to O_ anion-radical. On the other hant3, Cornazl> observed on unsupported Moos the de-
crease in intensity of [he signal due to D1o(V) by the adsorption of oxygen and did not observe any
new signah which agreed with the author's result and be proposed the doub]}• bonded model of 0,'-
ion. Such inconsistency may be based on the difference of the samples, that is, supported and unsup-
ported Moos. However, according to the author's experiments, oxygen adsorbs weakly and reversibly desorbs byevacuation at room temperature. These facts mean that oxygen adsorbs as Os molecule and
interacts weakly with [he substances. Therefore. Mo(V) will disappear by the electron transfer from
one Afo(V) or from the F-center to another Mo(\'), because of the increase in the instability of Mo(V)
20) P. Slahelin and P. Busch, Aely. Phys. rlcta, 23. 530 (1950) ZQ T. Kawaguchi,'Handotai no Kagaku; p. 28, >laruzen, Tokyo (1969)
21) A. F. Kip, C. Kitteh R. A. Levy and A. M. Ports, Phyt. Rev., 91, IOfi6 (1953) 23) \. \1'. Lord, ibid., 105. 756 (193i)
24) E. R. S. \Vinter, Advans. CaPoI., 10, L96 (1958) 25) P. Amignes aad 5. Teichaer, Dfrcuss. Faraday, Soc., 41, 362, 404 (1966)
26) C. \accache, Cheni, Plrys. Lef4, 11, 424 (1971) 27) J. H. Lunsford and J. P. Jayne, J. Chem. Phys, 44, 1481 (1966)
28) A. J. Tench and P. J. Holroyd, Chem, Commun.. 1968, 471. 29) ~*. B. lVoag aad J. H. Lunsford, J. Chem. Phyys., 56. 1664 (1972)
30) A. J. Tench. Trmit. Furad. Soc.. 68, 1t81 (1972) 3I) V. A. Shvl[s and V. B. Kazansky, J. Calal., 25, 113 (1972)
The Review of Physical Chemistry of Japan Vol. 46 No. 1 (1976)
8 N. Sotani
state induced by the adsorption of oxygen. It is well known that Oa molecule is paramagnetic~~l,
but adsorbed oxygen does not show any signal. Oa molecule will be inactive for ESR (taking :b=Q:
state). Mo(V) disappears as follows;
O, O,
~- -Mc (V)-O-Mo (V)- ~ -Mo (VI)-O-Mo (IV)-
or Oz ~z
T -Mo (V)-?-Mo(V)- ~ -MO (VI)-',f-Mo(VI)-
The disappearance of Mo(V) by Hs0 adsorption will be also interpreted as in the case of oxygen ad~
sorption.
Acknowledgments
The author is grateful to Professor Masatomo Hasegawa for his encouragement and helpiul advice.
Also the author is indebted to bSr. Nakano, Mr. Mamba, Mr. Pujii and il4r. Isagawa for their ESR
measurements and useful discussions.
College aJ Liberal Arls
Kobe Universily
Tsurakabxto Nada
Kobe 657
Japan
32) 33)
R, Beringer and J. G. Caitle Jr., P/rys. Rev., 81, 82 (1951) M, Tinkbm and ~l. W. P. Strnndberg, ibid.. 97, 937. 95] (1955)